03-28-2014 01:04 PM - edited 03-04-2019 10:40 PM
What are the Pros/Cons of the following two config examples?
My thought on it is they are actually the same but using different metrics to meet the "Target shaping rate". However, the Peak (be) packets are marked "1" and eligible for discard upstream. Thoughts?
Example 1:
policy-map QOS-OUT-POLICY
class VIDEO-IPP4-OUT
bandwidth percent 50
shape peak percent 25
-show policy-map interface "Output"
Serial0/0/0
Service-policy output: QOS-OUT-POLICY
Class-map: VIDEO-IPP4-OUT (match-any)
375854 packets, 377883039 bytes
5 minute offered rate 779000 bps, drop rate 2000 bps
Match: ip precedence 4
375854 packets, 377883039 bytes
5 minute rate 779000 bps
Queueing
queue limit 64 packets
(queue depth/total drops/no-buffer drops) 43/5774/0
(pkts output/bytes output) 370080/369779265
bandwidth 50% (768 kbps)
shape (peak) cir 384000, bc 3840, be 3840
target shape rate 768000
Example 2:
policy-map QOS-OUT-POLICY
class VIDEO-IPP4-OUT
bandwidth percent 50
shape average percent 50
-show policy-map interface "Output"
Service-policy output: QOS-OUT-POLICY
Class-map: VIDEO-IPP4-OUT (match-any)
107508 packets, 89205062 bytes
5 minute offered rate 239000 bps, drop rate 0 bps
Match: ip precedence 4
107508 packets, 89205062 bytes
5 minute rate 239000 bps
Queueing
queue limit 64 packets
(queue depth/total drops/no-buffer drops) 0/2262/0
(pkts output/bytes output) 105246/86646386
bandwidth 50% (768 kbps)
shape (average) cir 768000, bc 7680, be 7680
target shape rate 768000
Thanks in advance!
Solved! Go to Solution.
03-29-2014 05:40 PM
Hi,
When you use the shape AVERAGE command
your traffic for the class is limited to
your CIR (768000)
When you use the shape PEAK command your
traffic tries to burst above the CIR.
CIR x (1+BE/BC)
384000( 1+ 3840/3840)
384000 (1+1)
768000
But - Here the bursting traffic (Above 384000) is liable for
dropping in the event of conjestion.
Hope this helps
Regards
Alex
03-31-2014 06:57 AM
Disclaimer
The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.
Liability Disclaimer
In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.
Posting
On many Cisco platforms, using shape peak just results in 2x the transmission rate of shape average (for the same CIR).
On some Cisco documentation, shape peak is listed as having a reserve of additional bandwidth, that can be "borrowed" against, until it's used up. Unused "average" bandwidth "pays back" the borrowed bandwidth. I.e. the long term average should still be the CIR, but traffic would allowed to be more bursty (short term).
03-29-2014 05:40 PM
Hi,
When you use the shape AVERAGE command
your traffic for the class is limited to
your CIR (768000)
When you use the shape PEAK command your
traffic tries to burst above the CIR.
CIR x (1+BE/BC)
384000( 1+ 3840/3840)
384000 (1+1)
768000
But - Here the bursting traffic (Above 384000) is liable for
dropping in the event of conjestion.
Hope this helps
Regards
Alex
03-31-2014 06:57 AM
Disclaimer
The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.
Liability Disclaimer
In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.
Posting
On many Cisco platforms, using shape peak just results in 2x the transmission rate of shape average (for the same CIR).
On some Cisco documentation, shape peak is listed as having a reserve of additional bandwidth, that can be "borrowed" against, until it's used up. Unused "average" bandwidth "pays back" the borrowed bandwidth. I.e. the long term average should still be the CIR, but traffic would allowed to be more bursty (short term).
03-31-2014 08:38 AM
I agree with both answers and your answers coincides with what I was saying in my original discussion.
Scenario: Branch office with T1 circuit
Example 1 Config (Peak):
bandwidth percent 50 ===> 50% bandwidth reserved in congestion
shape peak percent 25 ===> cir of 384000 with Peak equaling bc 3840 + be 3840 (default) = 768k
Example 2 Config (Average):
bandwidth percent 50 ===> 50% bandwidth reserved in congestion
shape average percent 50 ===> cir of 768000 only uses bc 7680 = 768k
Synopis:
In both example configs they are meeting the requirement of not using more than 50% (768k) of the T1's available bandwidth. However, in the "Peak" configuration example, any packets over 384k is marked with tos bit of 1 making it eligible for discard or in other words the amount of bits over Bc that could be sent during Tc and subjected to drop eligibility. With the "Average" configuration, it uses (Bc) will only send packets upto the CIR.
Thanks for your replies and helping confirming the QOS features with me.
David Undernehr
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide